38. Intersubjectivity, Glitches, and Improvisation

This post is also available in: Catalan

Alessandro Duranti
University of California, Los Angeles


  1. Intersubjectivity as a condition of human existence

I borrow from phenomenology the idea that intersubjectivity is an a priori universal condition of our existence such that even though we might not be aware of the presence of other people or of the impact that their actions directly or indirectly have on us, we are always in “a world of others.” This has implications for all kinds of human activities as well as for perception. When we look at a table, we assume, without thinking, that this same table is perceptually accessible to others, even when they are not present. The durable presence of the table as a three-dimensional object is made possible by a shared but differentiated perceptual field where we can “trade places” with others, that is, we can imagine what it would be like to see the table from their point of view.

Intersubjectivity therefore is a precondition for interaction that allows for different degree of attunement. This happens, for instance, when we go from being passively aware of the people, sounds, or smells of the world we inhabit to actively attending to an aspect of the environment that has a strong attentional pull, e.g., a familiar song blasting out of the sound system of a store we just walked in.

  1. The interactive and emergent quality of meaningful action

An interpretation of intersubjectivity in terms of a shifting degree of attunement lends itself to an understanding of human action as unfolding in time. As such, intersubjective attunement adjusts to context-specific interactional demands. Some telling examples of how meaning unfolds in spontaneous speech production were provided almost forty years ago by Charles Goodwin in his book Conversational Organization, where he showed how a speaker is able to change the meaning of an utterance in the course of its production in order to adjust to a new recipient’s knowledge of past events.

In addition to showcasing the notion of “recipient design” introduced by Harvey Sacks in his lectures on conversation, Goodwin’s examples highlight the fact that speakers can sometimes quickly adjust the content of what they are saying without stopping and restarting their utterances. Speakers seem to be very apt at rethinking, rephrasing, and changing the meaning of what they might have been trying to say. As noted by the philosopher Maurice Merleau-Ponty, when speaking, we are not only expressing ourselves to others, we are also discovering our own thoughts.

The open-ended quality of communication and, more generally, action for phenomenologists like Merleau-Ponty is akin to the dialogical linguistics proposed by Mikhail M. Bakhtin for whom the meanings of utterances are not completely authored by their speakers because all utterances are populated by the intentions of earlier users. This implies that we can never be completely in control of the force or pragmatic effects of our utterances. Sometimes we might offend others even though we did not mean to or we might sound hesitant even though we feel resolute. The weight of past language uses can predetermine and limit our ability to use it for our own individual goals.

  1. A different concept of intersubjectivity: collective intentionality

The conceptualization of intersubjectivity as the basic condition of our sociality is by no means universally known or accepted. One competing model of intersubjectivity first proposed in the 1980s is the so-called “collective” or “shared” or “we-intentionality” model proposed by analytic philosophers. Their models, which are an extension of the causal models of individual agents, aim to explain how two or more people manage to perform actions that require cooperation like, for example, pushing a car, walking down the street, playing music, toasting each other, or having a conversation. For the shared intentionality model, all of these activities share something, namely, the intention or goal of the collective activity. But if, as mentioned above, even one person’s intentions might originally belong to others and speaking needs continuous adjustment, including changing the meaning of what we started out as saying, it would seem difficult to specify ahead of time exactly what the shared intention or goal of a joint activity is.

Take, for example, playing music together. This is one of the group activities mentioned by the shared intentionality proponents. Playing music requires a high level of cooperation (and coordination), but the identification of its “shared goal,” which is required by the collective or shared intentionality model, may turn out to be a challenge. Whereas in the case of musicians playing off a written or memorized piece one might assume (but not necessarily know) that the shared goal is the performance of the known score, the same could not be said of groups of people who are performing music together but not necessarily playing off an existing score, whether written or memorized. Philosopher and jazz musician Gary Hagberg, who evaluated the relevance of the concept of collective or shared intention for capturing group jazz improvisation, suggested that in this kind of joint activity, intentions are altered in the course of performance. Hagberg argued that when highly skilled jazz musicians play together, there is a level of creative work that goes beyond what any of the individuals involved could have “intended.” Even when jazz musicians can be said to improvise on a particular song, the original harmonic structure, rhythm, tempo or time signature may be altered to such an extent that the song they started from would no longer be recognizable by most people. It would thus be difficult to say exactly what goal they share except in some very vague terms like improvising or transforming a previously composed song into something else that is related to it but in unanticipated ways.

  1. Improvisation and the force of tradition

Someone might argue that jazz improvisation is not a good counter-example to the model of collective actions as sharing a goal because it is an artistic activity and as such is guided by what Jan Mukařovsky identified as the “aesthetic function,” which entails the violation of established norms. Are not most of our activities much more predictable than jazz collective improvisation? This is indeed an implicit view among linguists and anthropologists who have tended to pay more attention to rules and routine activities than to novel or creative activities. Even when Noam Chomsky in the 1960s recognized creativity as part of the language faculty, he kept competence and performance separate so that the temporal unfolding of utterance meanings was left out of his model. In anthropology, predictability has been at the core of the definition of “culture” as shared knowledge and of the notion of “ritual” as a sequence of more or less invariant acts, which are not originally authored by the speaker (see Roy Rappoport’s Ritual and Religion in the Making of Humanity). Especially in the 1970s, the concept of ritual as the performance of some culturally transmitted routine with a basic, essential form that could be adapted to particular contexts was applied to all kinds of communicative acts, from the most mundane activities like greetings to “traditional” oratory.

  1. The rare recognition of improvisation

In the late 1970s, two social scientists made direct reference to the role of improvisation in their writings: Pierre Bourdieu, whose notion of “habitus” as a set of “dispositions” included “orchestrated improvisation,” and Ruth Finnegan who, on the basis of her study of oratory and music performance, proposed a definition of improvisation as “composition-in-performance.” But the great majority of scholars in linguistics and anthropology did not engage with these ideas. Students of verbal art, myself included, while documenting variation across contexts continued to ignore the occurrence and need for improvisation even when speakers are relying on formulae and common patterns. In the 1980s, only a handful of ethnographers (e.g., Barbara Myerhoff and Renato Rosaldo) mentioned improvisation in their analyses. In my own case, it took several years of teaching together with the legendary jazz guitar player Kenny Burrell before I could reconceptualize what I had learned about Samoan oratory in terms of improvisation and a few more year to write a chapter about socialization into verbal improvisation in collaboration with Steven Black, a linguistic anthropologist with extensive experience in jazz music and performance.

  1. Improvisation as a hidden property of human action

Despite the increased focus in the social sciences on variation, variability, and diversity of all kinds, there has continued to be little empirical research directly addressing the role of improvisation in language, interaction, and culture. The exceptions are usually scholars who have had experience in one or more improvisational arts, like Howard Becker, Keith Sawyer, or Tim Ingold.

Paraphrasing Heraclitus’ famous saying Φύσις κρύπτεσθαι φιλεῖ, often translated as ‘Nature likes to hide,’ we could say that “improvisation likes to hide,” that is, it is hard to detect and be recognized. Even though the analysis of human interaction shows that people are quick to adjust to new circumstances, change course, and do something that could not have been planned or expected, the work that they do is hardly ever recognized as an act of improvisation.

One set of contexts where improvisation could be said “to show itself” is when participants find themselves in situations when things do not go as they had expected. This is particularly evident in public events when participants are following a protocol that requires shared knowledge and suddenly something does not work and the joint activity cannot be completed or needs to be repeated. These are situations in which improvisation is desperately needed and the audience might expect it. I have examined a number of these interactional troubles and called them “interactional glitches.” In addition to constituting a challenge to the shared intentionality model because they occur even when participants appear to have the right prerequisites for cooperation, these glitches also show that there are social constraints on the degree of improvisation that participants are allowed to exert. This suggests that improvisation is a mode of acting that is not equally distributed across individuals, activities, and situations. Thus, for example, I have found that in the context of public performance, the on-going close scrutiny of a large number of people may inhibit the use of the on-the-spot creativity that would otherwise result in a successful adjustment or repair.

  1. Improvisation as a condition for successful cooperation

Most of the time we believe that we know how to perform familiar tasks. We open the door of our car, sit inside, turn on the engine, and off we drive to work without much thought. When we arrive to our desk, we turn on our computer and, in a few seconds, we are reading and responding to messages. At lunchtime, we eat the snack we brought from home or walk to a nearby café or restaurant. These activities constitute our daily “rituals” and they are supported by verbal routines like greetings, polite formulas, or responses, which, in turn, are made up of a few items taken out of a vast repository of stock phrases that are available to us as speakers of a language. Just like jazz musicians can improvise melodic lines and rhythm patterns that are based on established conventions and a history of listening to other musicians, we can also move in the world and communicate successfully by adopting and adapting stock phrases that we have heard in the past or are being suggested by what someone else just said. Most of the time, we are given the right text to read, start at the right time, finish the toast before the music starts, and lift the glass at the right moment. When things go wrong, we are usually quick to retrace our steps, adjust, rephrase, or just wait until someone else saves us from embarrassment. All of this is made possible by the magic of improvisation as a force of both stability and creativity. The identification of such a force in human interaction suggests that cooperation in joint activities needs intersubjective attunement as well as flexibility, inventiveness, and tolerance for variation.